The Observer, October 13, 2006
Volume XXXIX, Issue 7
Case physics grad student
click to enlarge
To many students at Case, especially in physics, it is comforting to think that everything in the world can be explained through science. However, sometimes science proves that some things are really just a mystery.
Katherine Jones-Smith, a third-year graduate student in particle astrophysics at Case, will be presenting "Pollock's Paintings: Are They Really Fractal?" at the physics department colloquium next Thursday.
Jones-Smith's talk precedes publication of her paper in Nature disputing the idea of fractal analysis of Jackson Pollock's paintings promoted by Richard Taylor, a physicist at the University of Oregon.
Her work will disprove Taylor's assertion that Pollock's paintings can be considered fractal and defined by their fractal characteristics.
In 1999, Taylor published a paper in Nature titled "Fractal Analysis of Pollock's Drip Paintings." In the paper, Taylor argued that Pollock's paintings had similar fractal characteristics and that these characteristics helped explain why people had an affinity for the paintings. Taylor's paper was featured also in Physics World and Scientific American.
Jones-Smith discovered Taylor's paper in 2004 while searching for a topic on which to speak for a presentation.
To her, the idea of defining a Pollock painting by its fractal qualities seemed difficult, especially since the analysis took place on such a macroscopic level. Generally, fractal paintings have the ability to be viewed on a rather more microscopic level.
"I decided to make my own similar paintings in Photoshop and subject them to the same analysis," said Jones-Smith.
Her works could have been made by Pollock, according to her analysis. Jones-Smith chose to ignore this interesting discovery.
"It didn't seem relevant at the time," said Jones-Smith. "It was just a way for someone to determine why people enjoy Pollock's paintings." Recently, however, her work has become more relevant.
Half a dozen paintings have surfaced that may or may not be true Pollocks, and the New York Times asked Taylor to do a fractal analysis of the new paintings to determine whether or not they are real.
In February, the Times reported that Taylor's analysis showed that the new paintings did not have the same fractal characteristics as Pollock's other paintings, casting aspersions on the authenticity of the new paintings.
Jones-Smith almost immediately called her adviser, Harsh Mathur, and asked him to help her do some more research into her conclusion that Taylor's methods were unreliable.
Although her previous work had only focused on one set of criteria that Taylor had used in his analysis, she was confident that his other criteria also would not hold up.
Jones-Smith created more "paintings" of her own in Photoshop to prove her hypothesis, including a simple rendering of a field of hand-drawn stars.
"I made something that looks like a three-year-old drew it," said Jones-Smith.
However, this simple drawing fit all of Taylor's criteria for a Pollock painting, including having two distinct fractal dimensions within specific ranges.
"You can see that this is not, in fact, a Pollock painting, so it stands to reason that Taylor's analysis of the newly-found Pollocks is not useful," said Jones-Smith.
Jones-Smith's manuscript will be published in Nature within the next few weeks, in "Communications Arising," a section devoted to interesting or important comments and clarifications on material previously published in Nature.
Her Colloquium talk will take place nextThursday at 4:30 p.m. in Rockefeller 301, with coffee and cookies available to those who attend. Ellen Landau, an authority on the work of Pollock and a professor of art history at Case, will be presenting with Jones-Smith.
